How to Write a LinkedIn Recommendation With AI (and Ask for One)

To write a LinkedIn recommendation, start with how you know the person, describe one specific project they handled, name the result it produced, and close with a clear endorsement. Keep it to about 100 to 200 words. AI can turn your rough notes into a clean draft, but the specifics have to come from you.

This guide covers both sides: prompts that turn real work details into a credible recommendation for a colleague, and a request message that makes it easy for someone to write one for you. For other everyday tasks, see our copy-paste AI prompts for work.

What makes a LinkedIn recommendation worth reading?

Most recommendations fail the same way: they praise without proof. "Hardworking team player" could describe anyone, so recruiters skim past it. A useful recommendation reads like a short reference check, with evidence a stranger can believe.

Strong ones have five parts:

1. Relationship: how and how long you worked together.
2. Situation: the project or problem you saw up close.
3. Action: what they did, in plain words.
4. Result: what changed, with a number if you know one.
5. Endorsement: who should hire or work with them, and why.

Here is the difference specifics make:

What you need before you start

• A 1st-degree connection. LinkedIn's help page on recommendations says you can request them from 1st-degree connections you work with or have worked with, and giving one works the same way. Not connected yet? Send a connection request first.
• Three real facts: one project, one thing they did, one result. Use numbers only if you are sure of them.
• Their next goal. Skim their headline and About section so you stress the skills that matter for the role they want.
• An AI assistant. Any free plan works; compare options in our guide to the best AI assistant for work.

How to write a LinkedIn recommendation with AI in 6 steps

Step 1: Write down the raw facts

Before you open the AI tool, jot rough notes in your own words. For example: "Worked with Priya 2 years on the billing team. Rebuilt our refund process. Refund tickets fell from about 40 a week to 15. Calm with upset clients." Notes like these give the AI something true to work with.

Step 2: Paste them into a structured prompt

This is the most useful of the AI prompts for a LinkedIn recommendation, because it locks the AI to your facts. Replace each bracket with your own details:

Help me write a LinkedIn recommendation for [name]. I was their [relationship, for example peer on the marketing team] at [company] for [time period].

Use only these facts and add no others:
- Situation: [project or problem]
- What they did: [specific actions]
- Result: [what changed, numbers only if I gave them]
- A quality I saw firsthand: [short example]

Their next goal: [role they want].

Write 100 to 150 words in the first person, in plain US English. Open with how we worked together, end with a clear endorsement for their next role, and avoid cliches like "team player" or "rockstar."

The "use only these facts" line matters most. Without it, AI tools tend to fill gaps with confident claims nobody can back up.

Step 3: Ask the AI to cut the filler

First drafts lean on stock phrases. Run a second prompt in the same chat:

Review this recommendation. List every sentence that could describe almost anyone, then rewrite those sentences using only the facts I gave you. If a sentence has no supporting fact, delete it.

Step 4: Make it sound like you

Read the draft out loud. Swap words you would never say and add one detail only you would know. A paragraph that sounds nothing like you will feel off to people who know you.

Step 5: Fact-check every claim

Check dates, titles and numbers against what you know. Keep confidential details, such as client names or revenue figures, out of both the post and the AI tool. Unsure of a number? Describe the change without it.

Step 6: Post it on LinkedIn

Open their profile, select More, then Recommend. Choose your relationship and their position at the time, paste your text and send it. LinkedIn moves menu items between app updates, so check LinkedIn Help if a button has moved.

Expected result: they get your recommendation as a message and can accept it, dismiss it or ask you to revise it. Once accepted, it appears in the Recommendations section of their profile, as explained in LinkedIn's guide to accepting recommendations.

LinkedIn recommendation examples for a manager, peer and direct report

The relationship changes what readers want to hear. A direct report is someone who reports to you, while a peer works at your level. Use this table to decide what to stress:

Treat the samples below as models and swap in your own facts.

Example for a manager

I reported to Daniel for two years on a software company's customer success team. He inherited high turnover and fixed it with simple habits: weekly one-on-ones, clear Monday priorities and honest feedback given in private. Within a year we had stopped losing people, and three of us were promoted. Any team would be lucky to have him.

Example for a peer

Maya and I worked side by side as content marketers for 18 months. When blog traffic stalled, she audited every older article, merged duplicates and rewrote our 20 key pages, and organic signups grew over the next two quarters. Her edits are fast, kind and always make the piece clearer. I would recommend her for any senior content or SEO role.

Example for a direct report

I managed Sam for three years after he joined my operations team as a junior analyst. He automated our weekly inventory report, saving the team hours every Friday, and by his second year he was training new analysts. Sam asks sharp questions, owns his mistakes and follows through. He is ready to lead a team, and I would hire him again without hesitation.

Each pairs a concrete project with a visible result. If you manage people, our AI prompts for managers cover feedback and reviews in more depth.

How to ask for a LinkedIn recommendation

Make it easy to say yes: remind the person what you worked on together and suggest what they could mention, while leaving the final words to them.

1. Pick the right person: someone who saw your work directly and recently.
2. Gather two or three wins from past reviews or project notes. The ones you listed when you used AI to write your self-evaluation work well.
3. Send it through LinkedIn. Open their profile, select More, then Request a recommendation. Choose your relationship and your position at the time, then replace the default message with your own.
4. Offer to return the favor. It is a common courtesy.
5. Follow up once. If you hear nothing in a week or two, send one friendly reminder, then let it go.

A request message you can copy

Hi [name], I'm updating my LinkedIn profile as I look for [type of role], and I'd value a short recommendation from you, since you saw my work on [project] up close.

If it helps, you could mention:
- [win, for example cutting client onboarding from three weeks to one]
- [skill, for example running the weekly stakeholder update]
- [quality, for example staying calm during the system migration]

No pressure if you're short on time. I'm also happy to write one for you.

Thanks,
[your name]

To have AI personalize the message for each person, use this prompt:

Write a short, friendly LinkedIn message asking [name], my former [relationship], for a recommendation. Mention our work on [project] and suggest these points they could cover: [two or three wins]. Keep it under 120 words, make it easy to say no, and offer to write one for them in return.

If you have not spoken in a while, reconnect before you ask. Our LinkedIn message templates for networking include short notes for warming up an old contact.

For contacts who rarely use LinkedIn, ask by email first. These AI prompts for professional emails help you get the tone right.

Common mistakes to avoid

• Posting the first AI draft. Readers spot the smooth, interchangeable tone quickly. Edit until it sounds like you.
• Letting AI invent results. Every number and claim must be something you know is true, because your name is on it.
• Trading recommendations with people who never saw your work. It reads as hollow and can hurt your credibility.
• Treating it as a formal reference. Employers in the US and the UK usually still run their own reference checks (UK employers call the person who vouches for you a referee). A recommendation supports your profile but does not replace that.
• Ignoring company policy. Some employers limit what managers can say publicly about current or former staff, so check before recommending someone on your team.